Worm-Gear advice sought

[AjohnwParticipant @ajohnw51620]

If you can't google to see that an epicyclic gear needn't involve cutting internal gears why not just use a ladder.

John

–

[HopperParticipant @hopper]

Posted by Ajohnw on 14/05/2016 23:42:41:

If you can't google to see that an epicyclic gear needn't involve cutting internal gears why not just use a ladder.

John

–

Because his milling machine table is not long enough to make a ladder long enough.

[AjohnwParticipant @ajohnw51620]

Posted by Hopper on 15/05/2016 01:24:42:

Posted by Ajohnw on 14/05/2016 23:42:41:

If you can't google to see that an epicyclic gear needn't involve cutting internal gears why not just use a ladder.

John

–

Because his milling machine table is not long enough to make a ladder long enough.

Maybe make it in sections so it can fold up small as well.

John

–

[AjohnwParticipant @ajohnw51620]

Anyway the op needs to back step. If the worm is bigger than the drum winding the cord torque is reduced as far as the worm gear is concerned. Rather than get wrapped up in nm it's better to think levers. So of for instance the drum was the same diameter as the PCD of the worm wheel a 50kg load would put 50kg on the teeth and worm. If 1/2 the dia 25kg and so on.

That leaves efficiency which from memory is circa 30% suggesting that a safety factor of 3 might be a good idea.

Often when worms are used for power transmission twin start worms are used. I've never needed to find out why but suspect it's down to efficiency.

If this is a chandelier that currently is more of a thought experiment practicalities may be important. When the house was build did the builder put the ceiling rose close to a joist ? If not some floor boards may need to come up / the rose moved. I breathed a sigh of relief when my wife asked me to fit a rather heavy light fitting in a lounge as it was close. Not always so in other rooms. This house has lath and plaster which for very high loads would still be a problem. Some houses mount the roses directly on plaster board.

John

–

[Nicholas LeeParticipant @nicholaslee]

Dear John,

It is a newly built house, and a suitable joist was installed. Above the joist is just the roof tiles, so there's no gap to install a winch in the ceiling.

The Chandelier is in the middle of the ceiling, so there is nowhere to lean a ladder against, and a scaffold platform that high would cost circa £400 to hire each time, so a winch system was specified.

The title of this "Beginner's Questions" forum thread is "Worm-Gear advice sought"; so why be sarcastic about me not knowing to search for some variant of of an epicyclic gear that I have never heard of?

I was curious anyway, so I did try extensively Googling epicyclic gears, but all I found was several hundred examples of planetary gearboxes that all used a ring-gear with internal teeth.

Maybe it totally obvious to you how to make one without an internal ring gear, but it is not obvious to me.

I found one bizarre example that used planetary bevel gears, but that looked overly complicated.

More searching led me to "differential spur gear trains". Is that perhaps what you are alluding to?

Or, perhaps you meant a face-gear solution like this one?

All these solutions looks like an awful lot of gear cutting, to achieve what one worm gear stage can do.

Whilst it is an interesting subject, I think it might be academic for this application, given that we are lifting a load that requires the gearbox to be self-locking, and a worm-gearbox is best for that.

Regarding multi-start vs. single-start worms, I found this friendly website, where it says that:

"The worm drive inefficiency originates from the sliding contact between the teeth. A multi start thread has a steeper helix angle which results in less friction between the threads (so it is more efficient) and therefore such a system is less likely to be self-locking. It follows that a steeper helix allows for faster translation along the threads i.e. an item utilising a multi start thread can be tightened in fewer rotations than one using a single start thread."

So, multi-start threads are more efficient, but that can back-drive (which is bad for this application), and they only give half the gear ratio (which is also bad for this application).

Best Wishes

Nick Lee

 

Edited By Nicholas Lee on 15/05/2016 15:15:29

[Mark CParticipant @markc]

Nick,

I read the bit about making it yourself but have you tried searching for ready made units? I searched google quickly and there are lots available **LINK**

Mark

[John HaineParticipant @johnhaine32865]

Good grief! The one at that link has these characteristics:

• Max lifting load: 50kg (110 Pounds)
• …
• Lifting Speed: 1.3 Metres per Minute
• …
• Rated Power: 4KW
• Rated Current: 18A
• …

50 Kg is ~500 N weight. Lifting that at 1.3 m/min takes 1.3×500/60 = Joules/s = 7.5 watts. It must get mighty hot if its rated power is 4 kW, that's 5 hp!

[Michael BriggsParticipant @michaelbriggs82422]

Hello Nicholas,

I would consider getting Dell Boy and Rodney to do the job for you, I have heard their service is smashing.

Michael.

[Nicholas LeeParticipant @nicholaslee]

Hi,

Dell Boy and Rodney -LOL, I remember that sketch.

Well, £550 for a 50Kg winch is about an order of magnitude more than I want to pay, hence the DIY approach. Even buying commercially made gears would probably be in three-figures.

It is odd that the commercial winch is so power-hungry, as John Haine said, it uses a 4KW motor to do 7.5W of work. (This makes me nervous that I have forgotten some important calculation factor and my design is actually orders of magnitude under-powered)

Oddly, their website twice says what the 50Kg winch is 'efficient', despite physics disagreeing with their marketing department.

Curiously they also sell a differently designed hoist that can lift a whopping 450Kg, with only a 2.2KW motor. Perhaps their flat-disc shaped ones are particularly inefficient for some reason.

My earlier posted design parameters at 0.279m per minute = 0.00465m/s, moving with a force of 500N, implies a power of 2.325 Watts (at 100% efficiency)

The 919d1481 gearmotor I chose uses an RE540/1 motor, which is 61.9% efficient when run from 12V.

It has a power of 21.2 Watts. Of course, my proposed solution has 3 winches and so had 3 motors to help generate the required welly*.

Best Wishes

Nick Lee

*welly being the pre-SI unit of oomph

[Neil WyattModerator @neilwyatt]

Worth looking at Hopper's earlier posting. cheap pistol drills have epicylic gearboxes built in usually attached to a 540 motor.

As to the original unanswered question, you should do fine if you use an acme-style thread (30 degrees is close enough) and cut a hob in silver steel at the same setting as you make the worms, but a little deeper. Cut four or five flutes in the hob before hardening. Use a deeply gashed brass blank for each worm wheel and free-hobbing should be successful.

Neil

[JasonBModerator @jasonb]

With no lathe cutting the hob and making worms won't be easy! thats why I suggested using an off the shelf ACME tap to hob the wheel

Edited By JasonB on 15/05/2016 20:51:56

[John HaineParticipant @johnhaine32865]

I can't help thinking this is getting over-engineered Nicholas. In our church we have a number of wrought steel canelabra, I'm not sure how much they weigh but collectively must be ~50 Kg. The whole lot is suspended on one Kevlar cord, obtained from a chandlery amd IIRC 6mm diameter. As they hang catenary-fashion actually the tension in the cord is significantly more. They've been up for about 10 years without a problem. So for me an approach with one cord with 3 falls and pulleys seems obvious. You will be winding in no more cord than 3 separate winches. So – winch in the boss, cord goes up to the ceiling fitting, over a sheave, back down and through another sheave, back up and made fast. I don't think you would have any problem with uneven winding, and it will take no more power than 3 separate motors because of the mechanical advantage of the pulleys.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by JasonB on 15/05/2016 20:50:59:

With no lathe cutting the hob and making worms won't be easy! thats why I suggested using an off the shelf ACME tap to hob the wheel

.

… and [given that limitation] the hobbing process might be tricky on a 100 tooth wormwheel of the proposed diameter: That's one of the reasons why I suggested a cascade of low ratio reductions.

10:1 x 10:1 = 100:1

20 teeth on the wheel and a two-start worm might be neat.

MichaelG.

[Nicholas LeeParticipant @nicholaslee]

Hi,

To Neil:

An old 12V cordless drill will indeed have a nice DC motor and gearbox inside, but I'm not sure if it would generate enough torque by itself. It would need to be measured with a dynamometer to be sure. My gut feeling is that it would still need one extra stage of gearing down before it could safely drive the winch(es).

To Neil, & JasonB:

Only having a mill, I need to buy the worm, or make it from pre-threaded stock such as a Leadscrew.

For making the worm-wheel, Leadscrew taps are a bit pricey. Marchant Dice sell a Trapezoidal Leadscrew tap for £67.80, and yet the corresponding TR20x4D Leadscrew is only £11.76 per foot.

I should be able to save myself £56 by making myself a hob from a piece of the Leadscrew. I don't have any experience yet with hardening and annealing of hobs, but it will be good experience.

Admittedly this thread is Trapezoidal, rather than ACME or any of the other thread forms, but I'm not sure by how much that affects things.

Acme threads have 14 1/2 degree flank angles (29 degrees included) whereas Trapezoidal threads have 15 degree (30 degrees included) flank angles. I would imagine this difference will be insignificant.

NB: The only trouble with a multi-stage worm drive is that the axes turn 90-degrees with each stage, and my motor+worm drive need to be flat and in the same plane to avoid the chandelier boss from being too thick.

To JohnHaine:

I am quite tempted by your multi-pulley, single-winch, single-rope (Chandelier of Damocles) solution. I think I might even have a suitable chunkier 12V gearmotor that could drive it too. I just need to have a rummage. I think it was one of these. 12V, 5Nm, 47RPM, 27W

It would still need some extra gearing to get the required torque on the (single) capstan, but at least I would only then need one set instead of three, and I wouldn't need too much of a gear ratio to get what I need.

Best Wishes

Nick Lee

 

Edited By Nicholas Lee on 15/05/2016 21:48:29

[JasonBModerator @jasonb]

As the leadscrew is unlikely to be a carbon steel you won't easily harden it and I doubt case hardening (hard outer surface only) would be upto the job. The ME supplierts do chaeper taps, 5/8" x 10 or 8 tpi for £15

Regarding plugging in a 12v transformer into the light fitting and then use the mains 240V supply to the light to run the motors via a simple remote control.

At 10ft high can't a pair of steps be used?

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Nicholas Lee on 15/05/2016 21:45:42:

NB: The only trouble with a multi-stage worm drive is that the axes turn 90-degrees with each stage, and my motor+worm drive need to be flat and in the same plane to avoid the chandelier boss from being too thick.

.

Oddly enough … I saw that as a significant advantage

It would avoid the need for the 'axis-changing' pulleys.

MichaelG.

.

P.S.   I've just found this very informative catalogue cfrom Davall

Edited By Michael Gilligan on 16/05/2016 08:19:44

[John FieldingParticipant @johnfielding34086]

If I recall from my days of bell ringing, the main chandelier in the nave of St Paul's church in Durban has a ex Wellington bomber bomb winch mounted in the ringing chamber to lower the massive array when lamps needed to be changed. It is a hand wound contraption mounted on an ali bracket with lots of lightening holes, apparently salvaged from a crashed Wimpey at the end of WW2. Not sure what the chandelier weighed but the bomb winch handled up to 500lb bombs. It has a plate attached with its rating!

[Michael GilliganParticipant @michaelgilligan61133]

I've just had a thought [a dangerous occurrence, I know]

Given the infrequent use proposed for this device; might it be possible to make the wormwheels in the form of a 'lantern pinion' ? … with the trundles skewed at the helix-angle of the worm.

This would lend itself to manufacture on the Mill , and should be economical.

MichaelG.

[AjohnwParticipant @ajohnw51620]

Posted by Nicholas Lee on 15/05/2016 15:14:52:

Dear John,

It is a newly built house, and a suitable joist was installed. Above the joist is just the roof tiles, so there's no gap to install a winch in the ceiling.

The Chandelier is in the middle of the ceiling, so there is nowhere to lean a ladder against, and a scaffold platform that high would cost circa £400 to hire each time, so a winch system was specified.

The title of this "Beginner's Questions" forum thread is "Worm-Gear advice sought"; so why be sarcastic about me not knowing to search for some variant of of an epicyclic gear that I have never heard of?

I was curious anyway, so I did try extensively Googling epicyclic gears, but all I found was several hundred examples of planetary gearboxes that all used a ring-gear with internal teeth.

Maybe it totally obvious to you how to make one without an internal ring gear, but it is not obvious to me.

I found one bizarre example that used planetary bevel gears, but that looked overly complicated.

More searching led me to "differential spur gear trains". Is that perhaps what you are alluding to?

Or, perhaps you meant a face-gear solution like this one?

All these solutions looks like an awful lot of gear cutting, to achieve what one worm gear stage can do.

Whilst it is an interesting subject, I think it might be academic for this application, given that we are lifting a load that requires the gearbox to be self-locking, and a worm-gearbox is best for that.

Regarding multi-start vs. single-start worms, I found this friendly website, where it says that:

"The worm drive inefficiency originates from the sliding contact between the teeth. A multi start thread has a steeper helix angle which results in less friction between the threads (so it is more efficient) and therefore such a system is less likely to be self-locking. It follows that a steeper helix allows for faster translation along the threads i.e. an item utilising a multi start thread can be tightened in fewer rotations than one using a single start thread."

So, multi-start threads are more efficient, but that can back-drive (which is bad for this application), and they only give half the gear ratio (which is also bad for this application).

Best Wishes

Nick Lee

 

Edited By Nicholas Lee on 15/05/2016 15:15:29

I posted as shot of the arrangement in the other thread on epicyclic – I had time to take a shot when I posted in that one. The web page it's on has gone but the same info is available on the wayback machine here

**LINK**

While multistart worms increase efficiency I believe that twin start worms are often used in motor gearboxes to increase their load carrying capacity. I doubt of a low pitch 2 start worm would drive backwards but as you have no lathe you would need to buy a tap to hob a worm wheel – can't say as I have ever seen any 2 start taps.

I think Michael posted and idea that could be switched around a bit. A lantern wheel with a huge number of pins driven by a pinion. I suggest that the pinion would need to have 15 teeth minimum to get several acting at the same time. It could be hardened and tempered to increase it's strength. For the lantern wheel I would have thought that 3/16 dia silver steel as it comes would be fine for the pins providing it's supported at each end – sort of wheel with a groove around the periphery. Even 1/8 dia should do it providing there is very little unsupported length at each end. I'll leave stress calcs to some one else. The pitch would need to be greater than the dia suggests to provide clearance for simple teeth on the pinion. The flanks cut radially with a round or cycloid type end. This is the type of tooth form usually used on the gear rather than on the pinion.

The pins to go with that would have a dia of 0.9 x the gap between the teeth, They would mesh when an arc drawn from the centre of the pinion passed through the centre of the tooth rad and the centre of the pins. The motor could drive the pinion via a couple of commercial bevel gears.

John

–

PS Thanks for a copy of the pdf Keith.

Edited By Ajohnw on 16/05/2016 10:46:44

Edited By Ajohnw on 16/05/2016 10:49:02

[AjohnwParticipant @ajohnw51620]

That's windmill technology but the design is after Harrison's great wheels. The proportions should be correct but I can't be 100% sure of the depthing but that can be found after the parts are made.

John

–

[john carruthersParticipant @johncarruthers46255]

It used to be quite common for amteur telescope makers to produce their own worms and wheels.
Traditionally you make the worm leaving plenty to produce a hob at one end wich is later discarded.

**LINK**

[Author]

Posts